Nonmonotone Total Variation Minimization Based Projection Restoration For Low-Dose CT Reconstruction

Computed Tomography (CT) diagnosis is an important medical image diagnosis method like MRI, isotope scanning and ultrasound image to facilitate manual or computer-aids analysis. CT is one of the most active technologies in medical imaging because of its convenient, painless, high resolution, definite anatomy relationship and clear morbidity imaging. CT, in particular, can produce images from human body non-invasively that reveal the structure, function of internal organs or tissues. Currently, computed tomography technology has become an important method in tumour diagnosis and high-quality imaging of the coronary arteries. Some important information of CT images may be submerged due to the presence of noise. Moreover, as the new CT equipment used higher X-ray dose than conventional ones, making it more and more concerned about potential harm to humans from high X-ray dose. So, it is important for CT images to be preprocessed before being used for medical analysis and application, and minimizing x-ray exposure to the patients has been one of the major efforts in the CT fields.Recently, with the rapid development of hardware, a number of new algorithms and ideas appear from the field of image denoising and restoration, and this is provided new opportunities to improve the quality of low-dose CT images. Until now, a lot of common de-noise methods have been proposed, including median filter, wiener filter and histogram based filters. The quality of CT images is improved to some extent, and to satisfy some determinate applications. However, these methods do not solve the defects of the CT medical images radically. Presently wavelet threshold method and bilateral smoothing of image de-noising attract attention. But the former doesn't fit the images that have a low signal-to-noise ratio (SNR), and the latter has multi-parameters to adjust. Image de-noising based on the anisotropic diffusion filtering can preserve the edge of the image. However, the texture characteristics of CT images are weakened. As a result, it is difficult for doctor to diagnose different disease using blurred images.In order to improve the reconstruction quality of low-dose CT image, a new approach is proposed based on low-dose CT projection restoration in this paper. This method based on projection restoration combined the advantages of TV algorithm and the well known Barzilai-Borwein stepsize. First, projection data is transformed from Poisson distribution to Gaussian distribution using nonlinear Anscombe transform. Then, the Anscombe transformed data is filtered by an efficient nonmonotone total variation minimization denoising algorithm. Last, the reconstruction is achieved by inverse Anscombe transform and filtered back projection (FBP) method. Simulated and clinical low-dose CT data experimental results demonstrate that a high-quality CT image can be reconstructed.At first, the present and developmental state of research on CT image de-noising is introduced in this paper. Then, we describe the wide variety of medical image de-noising methods including wiener filter, anisotropic diffusion filtering, wavelet threshold method and bilateral smoothing and their applications. In chapter 4, we introduce the Nonmonotone Total Variation Minimization de-noising algorithm. Chapter 5 briefly summarizes the major contribution of the current work and provides some suggestions for the future research.